Undoubtedly at some stage Brownlee (or Pilot) changed from using the 12AH8 to the widely available ECH81. There is no evidence that the chassis was ever wired for the 12AH8.
The Pilot Mariner Mk II Radiogram chassis mainly differs from the radio only table model in having a ferrite rod and only one RF section coil on the chassis rather than three. This variation is also on the Trader sheet 1126 Pilot 85 series, though it uses two frame aerial coils.
The other difference is a third coil (2.2mH in air, 3mH on the rod) right on the very end of the ferrite rod connected across the aerial and earth terminals via chassis wiring. A slim coax cable also connects to this point to connnect to an aerial in the cabinet.
The IF appears to be 471 kHz, same as the UK Pilot 754, 85A and 85U. The T85M (R&TVS) is listed as 470 kHz.
ECH81 cathode: 2.2V, 220 Ohms, 47 K from triode grid to cathode.
6BA6 cathode: 0.75V, 62 Ohms.
The 6AT6 has feedback from the audio transformer secondary, like the Pilot "Blue Peter", with 33 Ohms cathode to chassis and 220 Ohms from speaker secondary. Only 15mV. The grid leak is 10M Ohms. 10nF connects the anode to the 6BW6 grid. A 270 pF mica capacitor on the anode decouples IF. Tone control (Bass, Music, Voice) is possibly via feedback. The same switch is used for normal and Bass Gram input.
The 6BW6 output differs from the Blue Peter and 85 series. The audio transformer has a tapped primary which feeds 2K (or possibly 2.5K, the paint was burnt off) to the dual 32uF HT capacitor. The centre tap is fed from the 1st 32uF and the 6X4 rectiier. Due to the feedback and type of audio transformer there is no "Tone Correction" / Snubber capacitor on the primary or 6BW6 anode. The cathode is 8.2V using a 270 Ohms decoupled with 25uF 25V rating.
6X4 HT: 253V DC (with 230V AC mains).
HT1 after 1st dropper fed from audio transformer: 180V. This feeds most ot the set. There is a 2K dropper at the 6BA6 decoupled with 100nF giving 82V.
Six paper dielelectric capacitors were replaced. Perhaps a couple not so critical:
2nF, 5nF, 10nF and 3 x 100nF.
The blank magic eye was replaced with a Russian NOS 6E5C (6e5s). One link is required for the target HT. There is a series 1M Ohm from AGC line to the eye grid, presumably to stop space charge affectiing the AGC voltage? The AGC from the final IF is decoupled by 1 M Ohms and 100nF. The feed to the switch selecting radio or gram to the 500K volume control uses a curious combination 47 K (IF stopper) and capacitor (less than 1nF?) that looks like a tubular ceramic capacitor with an extra leg at one end.
There is a thermistor on a tag of the transformer to chassis, use unknown. No obvious voltage selection.
The cabinet has a label near the turntable:
MODEL "MARINER AUTO-TRG."
6 VALVE TABLE RADIOGRAM
Though it has only five valves, including rectifier, on the chassis, the sixth is the Magic Eye.
The example chassis shows no evidence ot ever being wired for the 12AH8, the heater connections are for the ECH81. There is also possibly a wiring error on the example as there is no AGC connection to the ECH81 hexode g1. Actually the waveband selector was wired direct to the ECH81 even though the RF tuning capacitor is wired via a 5nF on the top of the chassis. This means as the switch is operated there is no bias on the ECH81 and at each position the g1 is shorted to 0V (chassis) by the LW or MW coil on the ferrite rod or the SW coil under the chassis. Thus the 5nF capacitor is redundant. All the examined UK Pilot schematics either isolate the coils via 2nF to 5nF and feed AGC via 1M Ohm, or there is no capacitor and then the "earthy" end of all the coils is decoupled and connected to the AGC. There is no way the earthy end of the coils in this set could connect to the AGC.
I don't remember the details of the Radio only version I restored.
There was a problem with changing gain and intermittent operation which was corrosion or tarnish on the wavechange switch. I changed the RF common of the switch direct to the RF tuning cap instead of the long wire under the chassis to ECH81 hexode g1. A 1M Ohm was added from the g1 to the nearby AGC line. AGC was observed to better compensate between strong and weak stations. Possibly there was a mistake made when changing from the 12AH8 version to the more logical ECH81 version in Pilot or Brownlee Dublin.
Tarrifs on imports is the likely explanation for the claimed Brownlee assembly.
The chassis has the same label as the radio only version. The rear of the chassis is quite far from the rear panel due to the depth of the speaker part and BSR turntable. Thus the slim coax soldered to the rear of the Aerial and Earth sockets likely connected to sockets on the missing rear panel. The gram and speaker sockets have the "earth" at the top. The chassis has a red and black dot beside the gram socket. The turntable power is via a 2 pin 5A plug with the inline style socket bolted to the rear of the radio chassis.
The autochanger has a flip over cartridge for 78/microgroove and speeds of 33, 45 and 78.
I checked the photos of the Radio only version. It has all the RF coils on the chassis and has no ferrite rod. The AGC connection can clearly be seen feeding the "earthy" end of all the RF coils, just like the UK Pilot T85M. It has no capacitor from the Hexode signal grid to the RF tuning capacitor which connects direct to the Wavechange switch.
The photo of the triode-hexode base is too obscured by parts and wires to be completely sure if the heater is wired for the 12AH8 or ECH81. It does look like pin 9 goes to chassis (12AH8), thus 4 and 5 would be shorted. In contrast the RG sample has a wire link on pin 5 to chassis (ECH81) and none is visible on the Radio only version. However it's clearly wired for AGC via the RF coils.
The Pilot EX854 and T854 have the coils all earthed, a capacitor isolating the 12AH8 hexode signal grid and AGC via 1M Ohm to the g1H.